[14072] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
| 3 | * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
| 4 | *
|
---|
| 5 | * This file is part of HeuristicLab.
|
---|
| 6 | *
|
---|
| 7 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
| 8 | * it under the terms of the GNU General Public License as published by
|
---|
| 9 | * the Free Software Foundation, either version 3 of the License, or
|
---|
| 10 | * (at your option) any later version.
|
---|
| 11 | *
|
---|
| 12 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 15 | * GNU General Public License for more details.
|
---|
| 16 | *
|
---|
| 17 | * You should have received a copy of the GNU General Public License
|
---|
| 18 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
| 19 | */
|
---|
| 20 | #endregion
|
---|
| 21 |
|
---|
| 22 | using System;
|
---|
| 23 | using System.Collections.Generic;
|
---|
[14302] | 24 | using System.Diagnostics;
|
---|
[14072] | 25 | using System.Linq;
|
---|
| 26 | using HeuristicLab.Common;
|
---|
| 27 | using HeuristicLab.Core;
|
---|
| 28 | using HeuristicLab.Data;
|
---|
| 29 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
| 30 | using HeuristicLab.Optimization;
|
---|
| 31 | using HeuristicLab.Parameters;
|
---|
| 32 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
| 33 |
|
---|
| 34 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression {
|
---|
| 35 | [Item("SymbolicRegressionSingleObjectiveOSGAEvaluator", "An evaluator which tries to predict when a child will not be able to fullfil offspring selection criteria, to save evaluation time.")]
|
---|
| 36 | [StorableClass]
|
---|
| 37 | public class SymbolicRegressionSingleObjectiveOsgaEvaluator : SymbolicRegressionSingleObjectiveEvaluator {
|
---|
| 38 | private const string RelativeParentChildQualityThresholdParameterName = "RelativeParentChildQualityThreshold";
|
---|
| 39 | private const string RelativeFitnessEvaluationIntervalSizeParameterName = "RelativeFitnessEvaluationIntervalSize";
|
---|
| 40 | private const string ResultCollectionParameterName = "Results";
|
---|
[14231] | 41 | private const string AggregateStatisticsParameterName = "AggregateStatistics";
|
---|
[14279] | 42 | private const string ActualSelectionPressureParameterName = "SelectionPressure";
|
---|
| 43 | private const string UseAdaptiveQualityThresholdParameterName = "UseAdaptiveQualityThreshold";
|
---|
| 44 | private const string UseFixedEvaluationIntervalsParameterName = "UseFixedEvaluationIntervals";
|
---|
[14072] | 45 |
|
---|
| 46 | #region parameters
|
---|
[14279] | 47 | public IFixedValueParameter<BoolValue> UseFixedEvaluationIntervalsParameter {
|
---|
| 48 | get { return (IFixedValueParameter<BoolValue>)Parameters[UseFixedEvaluationIntervalsParameterName]; }
|
---|
| 49 | }
|
---|
| 50 | public IFixedValueParameter<BoolValue> UseAdaptiveQualityThresholdParameter {
|
---|
| 51 | get { return (IFixedValueParameter<BoolValue>)Parameters[UseAdaptiveQualityThresholdParameterName]; }
|
---|
| 52 | }
|
---|
| 53 | public ILookupParameter<DoubleValue> ActualSelectionPressureParameter {
|
---|
| 54 | get { return (ILookupParameter<DoubleValue>)Parameters[ActualSelectionPressureParameterName]; }
|
---|
| 55 | }
|
---|
[14072] | 56 | public ILookupParameter<ResultCollection> ResultCollectionParameter {
|
---|
| 57 | get { return (ILookupParameter<ResultCollection>)Parameters[ResultCollectionParameterName]; }
|
---|
| 58 | }
|
---|
[14231] | 59 | public IValueParameter<BoolValue> AggregateStatisticsParameter {
|
---|
| 60 | get { return (IValueParameter<BoolValue>)Parameters[AggregateStatisticsParameterName]; }
|
---|
| 61 | }
|
---|
[14104] | 62 | public IValueParameter<IntMatrix> RejectedStatsParameter {
|
---|
| 63 | get { return (IValueParameter<IntMatrix>)Parameters["RejectedStats"]; }
|
---|
[14072] | 64 | }
|
---|
[14104] | 65 | public IValueParameter<IntMatrix> NotRejectedStatsParameter {
|
---|
| 66 | get { return (IValueParameter<IntMatrix>)Parameters["TotalStats"]; }
|
---|
[14072] | 67 | }
|
---|
| 68 | public IValueLookupParameter<DoubleValue> ComparisonFactorParameter {
|
---|
| 69 | get { return (ValueLookupParameter<DoubleValue>)Parameters["ComparisonFactor"]; }
|
---|
| 70 | }
|
---|
| 71 | public IFixedValueParameter<PercentValue> RelativeParentChildQualityThresholdParameter {
|
---|
| 72 | get { return (IFixedValueParameter<PercentValue>)Parameters[RelativeParentChildQualityThresholdParameterName]; }
|
---|
| 73 | }
|
---|
| 74 | public IFixedValueParameter<PercentValue> RelativeFitnessEvaluationIntervalSizeParameter {
|
---|
| 75 | get { return (IFixedValueParameter<PercentValue>)Parameters[RelativeFitnessEvaluationIntervalSizeParameterName]; }
|
---|
| 76 | }
|
---|
| 77 | public IScopeTreeLookupParameter<DoubleValue> ParentQualitiesParameter { get { return (IScopeTreeLookupParameter<DoubleValue>)Parameters["ParentQualities"]; } }
|
---|
| 78 | #endregion
|
---|
| 79 |
|
---|
| 80 | #region parameter properties
|
---|
[14279] | 81 | public bool UseFixedEvaluationIntervals {
|
---|
| 82 | get { return UseFixedEvaluationIntervalsParameter.Value.Value; }
|
---|
| 83 | set { UseFixedEvaluationIntervalsParameter.Value.Value = value; }
|
---|
| 84 | }
|
---|
| 85 | public bool UseAdaptiveQualityThreshold {
|
---|
| 86 | get { return UseAdaptiveQualityThresholdParameter.Value.Value; }
|
---|
| 87 | set { UseAdaptiveQualityThresholdParameter.Value.Value = value; }
|
---|
| 88 | }
|
---|
[14072] | 89 | public double RelativeParentChildQualityThreshold {
|
---|
| 90 | get { return RelativeParentChildQualityThresholdParameter.Value.Value; }
|
---|
| 91 | set { RelativeParentChildQualityThresholdParameter.Value.Value = value; }
|
---|
| 92 | }
|
---|
| 93 |
|
---|
| 94 | public double RelativeFitnessEvaluationIntervalSize {
|
---|
| 95 | get { return RelativeFitnessEvaluationIntervalSizeParameter.Value.Value; }
|
---|
| 96 | set { RelativeFitnessEvaluationIntervalSizeParameter.Value.Value = value; }
|
---|
| 97 | }
|
---|
| 98 |
|
---|
[14104] | 99 | public IntMatrix RejectedStats {
|
---|
| 100 | get { return RejectedStatsParameter.Value; }
|
---|
| 101 | set { RejectedStatsParameter.Value = value; }
|
---|
[14072] | 102 | }
|
---|
| 103 |
|
---|
[14104] | 104 | public IntMatrix TotalStats {
|
---|
| 105 | get { return NotRejectedStatsParameter.Value; }
|
---|
| 106 | set { NotRejectedStatsParameter.Value = value; }
|
---|
[14072] | 107 | }
|
---|
| 108 | #endregion
|
---|
| 109 |
|
---|
| 110 | public override bool Maximization {
|
---|
| 111 | get { return true; }
|
---|
| 112 | }
|
---|
| 113 |
|
---|
| 114 | public SymbolicRegressionSingleObjectiveOsgaEvaluator() {
|
---|
| 115 | Parameters.Add(new ValueLookupParameter<DoubleValue>("ComparisonFactor", "Determines if the quality should be compared to the better parent (1.0), to the worse (0.0) or to any linearly interpolated value between them."));
|
---|
[14104] | 116 | Parameters.Add(new FixedValueParameter<PercentValue>(RelativeParentChildQualityThresholdParameterName, new PercentValue(0.9)));
|
---|
[14072] | 117 | Parameters.Add(new FixedValueParameter<PercentValue>(RelativeFitnessEvaluationIntervalSizeParameterName, new PercentValue(0.1)));
|
---|
| 118 | Parameters.Add(new LookupParameter<ResultCollection>(ResultCollectionParameterName));
|
---|
| 119 | Parameters.Add(new ScopeTreeLookupParameter<DoubleValue>("ParentQualities") { ActualName = "Quality" });
|
---|
[14104] | 120 | Parameters.Add(new ValueParameter<IntMatrix>("RejectedStats", new IntMatrix()));
|
---|
| 121 | Parameters.Add(new ValueParameter<IntMatrix>("TotalStats", new IntMatrix()));
|
---|
[14231] | 122 | Parameters.Add(new ValueParameter<BoolValue>(AggregateStatisticsParameterName, new BoolValue(false)));
|
---|
[14279] | 123 | Parameters.Add(new LookupParameter<DoubleValue>(ActualSelectionPressureParameterName));
|
---|
| 124 | Parameters.Add(new FixedValueParameter<BoolValue>(UseAdaptiveQualityThresholdParameterName, new BoolValue(false)));
|
---|
| 125 | Parameters.Add(new FixedValueParameter<BoolValue>(UseFixedEvaluationIntervalsParameterName, new BoolValue(false)));
|
---|
[14072] | 126 | }
|
---|
| 127 |
|
---|
| 128 | [StorableHook(HookType.AfterDeserialization)]
|
---|
| 129 | private void AfterDeserialization() {
|
---|
[14279] | 130 | if (!Parameters.ContainsKey(ActualSelectionPressureParameterName))
|
---|
| 131 | Parameters.Add(new LookupParameter<DoubleValue>(ActualSelectionPressureParameterName));
|
---|
[14072] | 132 |
|
---|
[14279] | 133 | if (!Parameters.ContainsKey(UseAdaptiveQualityThresholdParameterName))
|
---|
| 134 | Parameters.Add(new FixedValueParameter<BoolValue>(UseAdaptiveQualityThresholdParameterName, new BoolValue(false)));
|
---|
[14104] | 135 |
|
---|
[14279] | 136 | if (!Parameters.ContainsKey(UseFixedEvaluationIntervalsParameterName))
|
---|
| 137 | Parameters.Add(new FixedValueParameter<BoolValue>(UseFixedEvaluationIntervalsParameterName, new BoolValue(false)));
|
---|
[14072] | 138 | }
|
---|
| 139 |
|
---|
| 140 | [StorableConstructor]
|
---|
| 141 | protected SymbolicRegressionSingleObjectiveOsgaEvaluator(bool deserializing) : base(deserializing) { }
|
---|
| 142 |
|
---|
| 143 | protected SymbolicRegressionSingleObjectiveOsgaEvaluator(SymbolicRegressionSingleObjectiveOsgaEvaluator original, Cloner cloner) : base(original, cloner) { }
|
---|
| 144 |
|
---|
| 145 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 146 | return new SymbolicRegressionSingleObjectiveOsgaEvaluator(this, cloner);
|
---|
| 147 | }
|
---|
| 148 |
|
---|
| 149 | public override void ClearState() {
|
---|
| 150 | base.ClearState();
|
---|
[14104] | 151 | RejectedStats = new IntMatrix();
|
---|
| 152 | TotalStats = new IntMatrix();
|
---|
[14072] | 153 | }
|
---|
| 154 |
|
---|
| 155 | public override IOperation InstrumentedApply() {
|
---|
| 156 | var solution = SymbolicExpressionTreeParameter.ActualValue;
|
---|
| 157 | IEnumerable<int> rows = GenerateRowsToEvaluate();
|
---|
| 158 |
|
---|
| 159 | var interpreter = SymbolicDataAnalysisTreeInterpreterParameter.ActualValue;
|
---|
| 160 | var estimationLimits = EstimationLimitsParameter.ActualValue;
|
---|
| 161 | var problemData = ProblemDataParameter.ActualValue;
|
---|
| 162 | var applyLinearScaling = ApplyLinearScalingParameter.ActualValue.Value;
|
---|
| 163 |
|
---|
| 164 | double quality;
|
---|
| 165 | var parentQualities = ParentQualitiesParameter.ActualValue;
|
---|
| 166 |
|
---|
| 167 | // parent subscopes are not present during evaluation of the initial population
|
---|
| 168 | if (parentQualities.Length > 0) {
|
---|
[14279] | 169 | quality = Calculate(interpreter, solution, estimationLimits, problemData, rows);
|
---|
[14072] | 170 | } else {
|
---|
| 171 | quality = Calculate(interpreter, solution, estimationLimits.Lower, estimationLimits.Upper, problemData, rows, applyLinearScaling);
|
---|
| 172 | }
|
---|
| 173 | QualityParameter.ActualValue = new DoubleValue(quality);
|
---|
| 174 |
|
---|
| 175 | return base.InstrumentedApply();
|
---|
| 176 | }
|
---|
| 177 |
|
---|
| 178 | public static double Calculate(ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, ISymbolicExpressionTree solution, double lowerEstimationLimit, double upperEstimationLimit, IRegressionProblemData problemData, IEnumerable<int> rows, bool applyLinearScaling) {
|
---|
| 179 | IEnumerable<double> estimatedValues = interpreter.GetSymbolicExpressionTreeValues(solution, problemData.Dataset, rows);
|
---|
| 180 | IEnumerable<double> targetValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, rows);
|
---|
| 181 | OnlineCalculatorError errorState;
|
---|
| 182 |
|
---|
| 183 | double r;
|
---|
| 184 | if (applyLinearScaling) {
|
---|
| 185 | var rCalculator = new OnlinePearsonsRCalculator();
|
---|
| 186 | CalculateWithScaling(targetValues, estimatedValues, lowerEstimationLimit, upperEstimationLimit, rCalculator, problemData.Dataset.Rows);
|
---|
| 187 | errorState = rCalculator.ErrorState;
|
---|
| 188 | r = rCalculator.R;
|
---|
| 189 | } else {
|
---|
| 190 | IEnumerable<double> boundedEstimatedValues = estimatedValues.LimitToRange(lowerEstimationLimit, upperEstimationLimit);
|
---|
| 191 | r = OnlinePearsonsRCalculator.Calculate(targetValues, boundedEstimatedValues, out errorState);
|
---|
| 192 | }
|
---|
| 193 | if (errorState != OnlineCalculatorError.None) return double.NaN;
|
---|
| 194 | return r * r;
|
---|
| 195 | }
|
---|
| 196 |
|
---|
[14279] | 197 | private double Calculate(ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, ISymbolicExpressionTree solution, DoubleLimit estimationLimits, IRegressionProblemData problemData, IEnumerable<int> rows) {
|
---|
[14302] | 198 | var estimatedValues = interpreter.GetSymbolicExpressionTreeValues(solution, problemData.Dataset, rows);
|
---|
| 199 | var targetValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, rows).ToList();
|
---|
[14072] | 200 |
|
---|
| 201 | var parentQualities = ParentQualitiesParameter.ActualValue.Select(x => x.Value);
|
---|
| 202 | var minQuality = parentQualities.Min();
|
---|
| 203 | var maxQuality = parentQualities.Max();
|
---|
| 204 | var comparisonFactor = ComparisonFactorParameter.ActualValue.Value;
|
---|
| 205 | var parentQuality = minQuality + (maxQuality - minQuality) * comparisonFactor;
|
---|
| 206 |
|
---|
| 207 |
|
---|
[14302] | 208 |
|
---|
[14280] | 209 | #region fixed intervals
|
---|
[14279] | 210 | if (UseFixedEvaluationIntervals) {
|
---|
[14302] | 211 | var e = estimatedValues.GetEnumerator();
|
---|
[14279] | 212 | double threshold = parentQuality * RelativeParentChildQualityThreshold;
|
---|
| 213 | if (UseAdaptiveQualityThreshold) {
|
---|
| 214 | var actualSelectionPressure = ActualSelectionPressureParameter.ActualValue;
|
---|
| 215 | if (actualSelectionPressure != null)
|
---|
| 216 | threshold = parentQuality * (1 - actualSelectionPressure.Value / 100.0);
|
---|
[14072] | 217 | }
|
---|
[14231] | 218 |
|
---|
[14279] | 219 | var pearsonRCalculator = new OnlinePearsonsRCalculator();
|
---|
| 220 | var targetValuesEnumerator = targetValues.GetEnumerator();
|
---|
| 221 | var trainingPartitionSize = problemData.TrainingPartition.Size;
|
---|
| 222 | var interval = (int)Math.Floor(trainingPartitionSize * RelativeFitnessEvaluationIntervalSize);
|
---|
[14231] | 223 |
|
---|
[14279] | 224 | var aggregateStatistics = AggregateStatisticsParameter.Value.Value;
|
---|
| 225 | var i = 0;
|
---|
| 226 | if (aggregateStatistics) {
|
---|
| 227 | var trainingEnd = problemData.TrainingPartition.End;
|
---|
| 228 | var qualityPerInterval = new List<double>();
|
---|
| 229 | while (targetValuesEnumerator.MoveNext() && e.MoveNext()) {
|
---|
| 230 | pearsonRCalculator.Add(targetValuesEnumerator.Current, e.Current);
|
---|
| 231 | ++i;
|
---|
| 232 | if (i % interval == 0 || i == trainingPartitionSize) {
|
---|
| 233 | var q = pearsonRCalculator.ErrorState != OnlineCalculatorError.None ? double.NaN : pearsonRCalculator.R;
|
---|
| 234 | qualityPerInterval.Add(q * q);
|
---|
| 235 | }
|
---|
[14231] | 236 | }
|
---|
[14279] | 237 | var r = pearsonRCalculator.ErrorState != OnlineCalculatorError.None ? double.NaN : pearsonRCalculator.R;
|
---|
| 238 | var actualQuality = r * r;
|
---|
[14231] | 239 |
|
---|
[14279] | 240 | bool predictedRejected = false;
|
---|
[14231] | 241 |
|
---|
[14279] | 242 | i = 0;
|
---|
| 243 | double quality = actualQuality;
|
---|
| 244 | foreach (var q in qualityPerInterval) {
|
---|
| 245 | if (double.IsNaN(q) || !(q > threshold)) {
|
---|
| 246 | predictedRejected = true;
|
---|
| 247 | quality = q;
|
---|
| 248 | break;
|
---|
| 249 | }
|
---|
| 250 | ++i;
|
---|
| 251 | }
|
---|
| 252 |
|
---|
| 253 | var actuallyRejected = !(actualQuality > parentQuality);
|
---|
| 254 |
|
---|
| 255 | if (RejectedStats.Rows == 0 || TotalStats.Rows == 0) {
|
---|
| 256 | RejectedStats = new IntMatrix(2, qualityPerInterval.Count);
|
---|
| 257 | RejectedStats.RowNames = new[] { "Predicted", "Actual" };
|
---|
| 258 | RejectedStats.ColumnNames = Enumerable.Range(1, RejectedStats.Columns).Select(x => string.Format("0-{0}", Math.Min(trainingEnd, x * interval)));
|
---|
| 259 | TotalStats = new IntMatrix(2, 2);
|
---|
| 260 | TotalStats.RowNames = new[] { "Predicted", "Actual" };
|
---|
| 261 | TotalStats.ColumnNames = new[] { "Rejected", "Not Rejected" };
|
---|
| 262 | }
|
---|
| 263 | // gather some statistics
|
---|
| 264 | if (predictedRejected) {
|
---|
| 265 | RejectedStats[0, i]++;
|
---|
| 266 | TotalStats[0, 0]++;
|
---|
| 267 | } else {
|
---|
| 268 | TotalStats[0, 1]++;
|
---|
| 269 | }
|
---|
| 270 | if (actuallyRejected) {
|
---|
| 271 | TotalStats[1, 0]++;
|
---|
| 272 | } else {
|
---|
| 273 | TotalStats[1, 1]++;
|
---|
| 274 | }
|
---|
| 275 | if (predictedRejected && actuallyRejected) {
|
---|
| 276 | RejectedStats[1, i]++;
|
---|
| 277 | }
|
---|
| 278 | return quality;
|
---|
[14231] | 279 | } else {
|
---|
[14279] | 280 | while (targetValuesEnumerator.MoveNext() && e.MoveNext()) {
|
---|
| 281 | pearsonRCalculator.Add(targetValuesEnumerator.Current, e.Current);
|
---|
| 282 | ++i;
|
---|
| 283 | if (i % interval == 0 || i == trainingPartitionSize) {
|
---|
| 284 | var q = pearsonRCalculator.ErrorState != OnlineCalculatorError.None ? double.NaN : pearsonRCalculator.R;
|
---|
| 285 | var quality = q * q;
|
---|
| 286 | if (!(quality > threshold))
|
---|
| 287 | return quality;
|
---|
| 288 | }
|
---|
| 289 | }
|
---|
| 290 | var r = pearsonRCalculator.ErrorState != OnlineCalculatorError.None ? double.NaN : pearsonRCalculator.R;
|
---|
| 291 | var actualQuality = r * r;
|
---|
| 292 | return actualQuality;
|
---|
[14231] | 293 | }
|
---|
[14302] | 294 | #endregion
|
---|
[14231] | 295 | } else {
|
---|
[14301] | 296 | var lsc = new OnlineLinearScalingParameterCalculator();
|
---|
| 297 | var rcalc = new OnlinePearsonsRCalculator();
|
---|
[14302] | 298 | var actualQuality = SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator.Calculate(SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, SymbolicExpressionTreeParameter.ActualValue, estimationLimits.Lower, estimationLimits.Upper, problemData, rows, true);
|
---|
| 299 |
|
---|
| 300 | var values = estimatedValues.Zip(targetValues, (es, t) => new { Estimated = es, Target = t });
|
---|
| 301 | int calculatedRows = 0;
|
---|
| 302 | double quality = 0.0;
|
---|
| 303 |
|
---|
| 304 | foreach (var value in values) {
|
---|
| 305 | lsc.Add(value.Estimated, value.Target);
|
---|
| 306 | rcalc.Add(value.Estimated, value.Target);
|
---|
| 307 | calculatedRows++;
|
---|
| 308 |
|
---|
| 309 | if (calculatedRows % 5 == 0) {
|
---|
| 310 | var alpha = lsc.Alpha;
|
---|
| 311 | var beta = lsc.Beta;
|
---|
| 312 |
|
---|
| 313 | OnlinePearsonsRCalculator calc = (OnlinePearsonsRCalculator)rcalc.Clone();
|
---|
| 314 | foreach (var t in targetValues.Skip(calculatedRows)) {
|
---|
| 315 | var scaledTarget = (t - alpha) / beta;
|
---|
| 316 | calc.Add(scaledTarget, t);
|
---|
| 317 | }
|
---|
| 318 |
|
---|
| 319 | var r = calc.ErrorState == OnlineCalculatorError.None ? calc.R : double.NaN;
|
---|
| 320 | quality = r * r;
|
---|
| 321 |
|
---|
| 322 | if (quality < parentQuality && actualQuality > parentQuality) {
|
---|
| 323 | Debugger.Break();
|
---|
| 324 | }
|
---|
| 325 | if (quality < parentQuality) return quality;
|
---|
[14231] | 326 | }
|
---|
[14302] | 327 | }
|
---|
| 328 |
|
---|
| 329 | //calculate quality for all rows
|
---|
| 330 | {
|
---|
| 331 | var r = rcalc.ErrorState == OnlineCalculatorError.None ? rcalc.R : double.NaN;
|
---|
[14279] | 332 | quality = r * r;
|
---|
[14302] | 333 | if (quality < parentQuality && actualQuality > parentQuality) {
|
---|
| 334 | Debugger.Break();
|
---|
[14301] | 335 | }
|
---|
[14302] | 336 | if (double.IsNaN(quality)) quality = 0.0;
|
---|
| 337 | if (quality != actualQuality) Debugger.Break();
|
---|
| 338 |
|
---|
| 339 | //necessary due to rounding errors and diff in the range of 10E-8
|
---|
| 340 | quality = actualQuality;
|
---|
[14231] | 341 | }
|
---|
[14301] | 342 |
|
---|
[14279] | 343 | return quality;
|
---|
[14072] | 344 | }
|
---|
| 345 | }
|
---|
| 346 |
|
---|
| 347 | public override double Evaluate(IExecutionContext context, ISymbolicExpressionTree tree, IRegressionProblemData problemData, IEnumerable<int> rows) {
|
---|
| 348 | SymbolicDataAnalysisTreeInterpreterParameter.ExecutionContext = context;
|
---|
| 349 | EstimationLimitsParameter.ExecutionContext = context;
|
---|
| 350 | ApplyLinearScalingParameter.ExecutionContext = context;
|
---|
| 351 |
|
---|
| 352 | var interpreter = SymbolicDataAnalysisTreeInterpreterParameter.ActualValue;
|
---|
| 353 | var estimationLimits = EstimationLimitsParameter.ActualValue;
|
---|
| 354 | var applyLinearScaling = ApplyLinearScalingParameter.ActualValue.Value;
|
---|
| 355 |
|
---|
| 356 | double r2 = Calculate(interpreter, tree, estimationLimits.Lower, estimationLimits.Upper, problemData, rows, applyLinearScaling);
|
---|
| 357 |
|
---|
| 358 | SymbolicDataAnalysisTreeInterpreterParameter.ExecutionContext = null;
|
---|
| 359 | EstimationLimitsParameter.ExecutionContext = null;
|
---|
| 360 | ApplyLinearScalingParameter.ExecutionContext = null;
|
---|
| 361 |
|
---|
| 362 | return r2;
|
---|
| 363 | }
|
---|
| 364 | }
|
---|
| 365 | }
|
---|